Characterizing the use of perdeuteration in NMR studies of large proteins C-13, N-15 and H-1 assignments of human carbonic anhydrase II

Perdeuteration of all non-exchangeable proton sites can significantly increase the size of proteins and protein complexes for which NMR resonance assignments and structural studies are possible. Backbone H-1, N-15, (CO)-C-13, C-13(alpha) and C-13(beta) chemical shifts and aliphatic side-chain C-13 and H-1(N)/N-15 chemical shifts for human carbonic anhydrase II (HCA II), a 259 residue 29 kDa metalloenzyme, have been determined using a strategy based on 2D, 3D and 4D heteronuclear NMR experiments, and on perdeuterated C-13/N-15-labeled protein. To date, HCA II is one of the largest monomeric proteins studied in detail by high-resolution NMR. Of the backbone resonances, 85% have been assigned using fully protonated N-15 and C-3/N-15-labeled protein in conjunction with established procedures based on now standard 2D and 3D NMR experiments. HCA II has been perdeuterated both to complete the backbone resonance assignment and to assign the aliphatic side-chain C-13 and H-1(N)/N-15 resonances. The incorporation of H-2 into HCA II dramatically decreases the rate of C-13 and (HNT2)-H-1 relaxation. This, in turn, increases the sensitivity of several key H-1/C-13/N-15 triple-resonance correlation experiments. Many otherwise marginal heteronuclear 3D and 4D correlation experiments, which are important to the assignment strategy detailed herein, can now be executed successfully on HCA II. Further analysis suggests that, from the perspective of sensitivity, perdeuteration should allow other proteins with rotational correlation times significantly longer than HCA II (tau(c) = 11.4 ns) to be studied successfully with these experiments. Two different protocols have been used to characterize the secondary structure of HCA II from backbone chemical-shift data. Secondary structural elements determined in this manner compare favorably with those elements determined from a consensus analysis of the HCA II crystal structure. Finally, having outlined a general strategy for assigning backbone and side-chain resonances in a perdeuterated large protein, we propose a strategy whereby this information can be used to glean more detailed structural information from the partially or fully protonated protein equivalent. (C) 1996 Academic Press Limited